From our preliminary studies, in which we followed spatial-temporal distribution of cell proliferation through the course of radiation treatment, tumor heterogeneity was detected before the treatment. In addition, a significant re-distribution of the proliferative capacity was observed in the tumor during the treatment, including significant and unexpected accelerated repopulation. Advances in radiation treatment delivery, such as intensity-modulated radiotherapy, provide the ability to customize radiation delivery based on physical conformity. Incorporation of regional biologic information, especially its changes during therapy, could lead to treatment adaptation. Similarly, early assessment of chemotherapy efficacy and lead to improved cancer management. The spatio-temporal response represents significant challenges for image-guided interventions; and the potential for significant improvement of the current treatment practice. Quantification of the changes and better understanding of tumor kinetics are the main unknowns preventing treatment adaptation. Our research will be focused in these two areas. Our main goals in this project are: 1) to estimate inter- and intra-tumor heterogeneity of cell proliferation and hypoxia for two tumor types and quantify its spatio-temporal change during cancer therapy and 2) to develop and test the methodology for quantitative monitoring of spatio- temporal cell proliferation and hypoxia distribution in the tumor and surrounding critical structures during cancer therapy. FLT will be used as a proliferation marker and Cu-ATSM as a hypoxia marker for in-vivo PET imaging. The PET imaging results will be correlated to immunohistochemical analyses of selected biopsy samples to further validate the used radiopharmaceuticals. Advanced image co-registration techniques (e.g., rigid body, deformable registration) and spatial statistics methods (e.g., Moran, G statistics) will be developed and applied to establish appropriate level of image and data analyses. Our methodology will be developed on larger animal tumor models (dogs), which are an excellent research model due to its large size and biological similarity to humans. Two spontaneous dog tumor types - soft tissue sarcoma (treated with radiotherapy) and non-Hodgkin's lymphoma (treated with chemotherapy) will be investigated. In addition to proving J feasibility, the data collected in this pilot study will be used to perform well-grounded power calculations for a future larger study and translation to humans. By combining concurrent monitoring of proliferation and hypoxia (FLT/Cu-ATSM) distributions and advanced image analyses, we hypothesize that spatio-al temporal quantification of the magnitude, and the change of both parameters, can provide quantitative basis for molecular image guidance of cancer therapy. [unreadable] [unreadable]